Multiple introductions boosted genetic diversity in the invasive range
of black cherry (Prunus serotina; Rosaceae)
Marie Pairon1, Blaise Petitpierre2, Michael Campbell3, Antoine Guisan2, Olivier Broennimann2, Philippe V. Baret1, Anne-Laure Jacquemart1,* and Guillaume Besnard2,†
1Earth and Life Institute, Research group ‘genetics, reproduction, populations’, Universite´ catholique de Louvain, Croix du Sud,
2 box 14, 1348 Louvain-la-Neuve, Belgium, 2Department of Ecology and Evolution, Biophore, University of Lausanne, 1015 Lausanne, Switzerland and3Penn State Erie, The Behredn College, School of Science, 4205 College Drive, Erie, PA 16563, USA
* For correspondence: E-mail [email protected]
†Present address: Imperial College, Silwood Park Campus, Buckhurst Road, Ascot, Berkshire SL5 7PY, UK
Received: 30 November 2009 Returned for revision: 29 January 2010 Accepted: 15 February 2010 Published electronically: 15 April 2010
† Background and Aims Black cherry (Prunus serotina) is a North American tree that is rapidly invading European forests. This species was introduced first as an ornamental plant then it was massively planted by for-esters in many countries but its origins and the process of invasion remain poorly documented. Based on a genetic survey of both native and invasive ranges, the invasion history of black cherry was investigated by iden-tifying putative source populations and then assessing the importance of multiple introductions on the mainten-ance of gene diversity.
† Methods Genetic variability and structure of 23 populations from the invasive range and 22 populations from the native range were analysed using eight nuclear microsatellite loci and five chloroplast DNA regions. † Key Results Chloroplast DNA diversity suggests there were multiple introductions from a single geographic region (the north-eastern United States). A low reduction of genetic diversity was observed in the invasive range for both nuclear and plastid genomes. High propagule pressure including both the size and number of intro-ductions shaped the genetic structure in Europe and boosted genetic diversity. Populations from Denmark, The Netherlands, Belgium and Germany showed high genetic diversity and low differentiation among populations, supporting the hypothesis that numerous introduction events, including multiple individuals and exchanges between sites, have taken place during two centuries of plantation.
† Conclusions This study postulates that the invasive black cherry has originated from east of the Appalachian Mountains (mainly the Allegheny plateau) and its invasiveness in north-western Europe is mainly due to multiple introductions containing high numbers of individuals.
Key words:Microsatellites, putative sources, invasive tree species, phylogeography, population genetics, black cherry, Prunus serotina var. serotina.
I N T RO D U C T I O N
The success of non-indigenous plants in areas distant from their native range have caused considerable economic and ecological costs worldwide (e.g. Wilcove et al., 1998; Forman, 2003; Simberloff, 2003; Pimentel et al., 2005). Different stages are usually recognized in the invasion process, ranging from the introduction of a species to its spread into the new range (With, 2002;Lockwood et al., 2005;Theoharides and Dukes, 2007). Because invasive species are introduced by human activities either intentionally or unintentionally, introduction modes can greatly vary among species. Propagule pressure con-stitutes a particularly important factor in the first steps of inva-sion. This composite term accounts for both the number of individuals – propagule size – within a single introduction event and the number of independent introductions – propagule number (Lockwood et al., 2005). Both propagule size and number play a major role in the dynamics of species’ establish-ment. The introduction of a small number of individuals in a new environment can potentially limit the establishment of a species if there is not enough genetic variation for the adaptive
evolutionary changes required by the new selection pressures to take place (Suarez and Tsutsui, 2008;Okada et al., 2009). On the other hand, multiple introductions and subsequent gene flow between populations may maintain part or all genetic vari-ation present in native populvari-ations (Carlton, 1996;Novak and Mack, 2001; Lockwood et al., 2005; Facon et al., 2008; Prentis et al., 2008;Henry et al., 2009; Wilson et al., 2009). The influence of these multiple introductions has not yet been fully resolved and a recent review has pointed out that the influ-ence of propagule number on the ability of a species to spread is dependent on both spatial and temporal aspects of the multiple introductions (Dlugosch and Parker, 2008). Nevertheless, mul-tiple introductions from genetically differentiated source popu-lations have been increasingly found in successful plant invasions (Hufbauer and Sforza, 2008; Prentis et al., 2009). The time needed for populations resulting from different intro-ductions to become connected is then dependent on the initial size of the introduced populations and the geographical distance separating them.
Key steps in the understanding and management of biologi-cal invasions are the frequency with which a species is
# The Author 2010. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For Permissions, please email: [email protected]
introduced into a specific area, the size of each introduction and the subsequent pattern of spread across the new range. Several population genetic studies have been devoted to answer these questions by identifying source populations (e.g. Carlton, 1996; Meekins et al., 2001; Besnard et al., 2007; Schlaepfer et al., 2008), comparing genetic structure of invasive populations with those in the native range (e.g. Eckert et al., 1996;Prentis et al., 2009), unravelling invasion histories (e.g. Carlton, 1996; Neuffer and Hurka, 1999; Meekins et al., 2001; Gaskin et al., 2005; Le Roux et al., 2008;Henry et al., 2009), or seeking the role played by evol-utionary processes in establishment and spread (for reviews, see Prentis et al., 2008; Suarez and Tsutsui, 2008).
Combining molecular markers with different modes of inheritance can be highly useful to assess population structure in invasive species (e.g.Gaskin et al., 2005;Grapputo et al., 2005). Chloroplast DNA (cpDNA) is generally maternally inherited in angiosperms, and cpDNA markers thus record gene flow coming from seed movement only, whereas nuclear markers record gene flow of both pollen and seeds (Devos et al., 2003). In addition, polymorphisms in haploid genomes are more affected by genetic drift than those from the nuclear genome (Petit et al., 1993). Patterns of the spatial distribution of cpDNA polymorphism established by seed dispersal during range expansion should therefore be more slowly eroded by subsequent gene flow, when compared with spatial patterns in nuclear genetic markers, especially in trees (e.g.McCauley et al., 2003;Petit and Hampe, 2006).
Trees have long been transferred across countries and conti-nents without being considered as potential threats (Petit et al., 2004;Bucharova and van Kleunen, 2009). In most cases, inva-sive trees have a long history of succesinva-sive introductions before being eventually classified as problematic aliens. Among these trees is the black cherry (Prunus serotina; Rosaceae), which was one of the first North American tree species to be introduced into Europe (Starfinger et al., 2003). Five varieties are usually recognized in its natural range that covers most of eastern north America (i.e. var. serotina, eximia, rufula, virens and salicifolia; McVaugh, 1951). The most common and widespread variety serotina produces valued furniture wood, but only in a limited portion of its range on the Allegheny plateau of Pennsylvania, New York and West Virginia (Marquis, 1990). The first record of the species in Europe dates back to 1623 when the tree was planted for ornamental purposes near Paris (Starfinger, 1997). Until the late 18th century, P. serotina was sparsely planted in parks and gardens in several European countries. Recommended in 1810 for sylvicultural practices, it was then more intensively planted for growing high quality timber trees and increasing forest production mainly on poor sandy soils (Muys et al., 1992). These experimental plantations rarely produced good quality wood. Between 1900 and 1930, black cherry was planted for multiple uses such as wind and firebreaks, improving the sandy soils under coniferous planta-tions, or providing shelter (Starfinger, 1997;Starfinger et al., 2003; Pairon et al., 2006a; Verheyen et al., 2007; Vanhellemont et al., 2010). Massive underplanting and filling of conifer stands occurred until the late fifties in Belgium, Germany and The Netherlands (Muys et al., 1992). Black cherry has been planted for forestry purposes since the
1980s in Poland and Romania and local introductions were performed in Italy and the UK (Muys et al., 1992). Presently, the species grows spontaneously in many European countries from France to Poland and Romania as well as from Denmark to Italy. Prunus serotina is spreading throughout temperate forests in north-western Europe, especially on well-drained poor soils. This tree is now con-sidered as one of the 100 most invasive alien species in Europe (DAISIE, 2009). It competes for resources with native plant species, especially during forest regeneration and under high herbivore pressure (e.g. Verheyen et al., 2007; Chabrerie et al., 2008; Vanhellemont et al., 2009). Invaded stands have higher levels of phosphorus, a shallower litter layer, and lower pH values than non-invaded stands (Chabrerie et al., 2008). Shading out light-demanding species, especially seedlings of other tree species, it can impede natural regeneration of native tree species and induces a decrease in species richness, mainly in disturbed stands. Fruits are produced in high quantities and are well persed, mainly by birds. More than 50 % of the seeds are dis-persed at higher distances than 50 m; the average dispersal distance being estimated to 257 m (Pairon, 2007; Chabrerie et al., 2008). Despite important investment, the control of its spread remains ineffective (Vanhellemont et al., 2009).
In this study, the invasion history of black cherry in Europe was inferred from genetic patterns. Both cpDNA markers and nuclear microsatellite loci were used to genotype native and invasive populations. This study was undertaken to (a) find the putative source(s) of European populations and (b) test whether invasive populations have lower genetic diversity than native populations.
M AT E R I A L S A N D M E T H O D S
Sampling scheme
Prunus serotina Ehrh. var. serotina plant material was col-lected from 23 populations in the invasive range and from 22 populations in the native range (Table 1) for a total number of 442 and 321 trees, respectively. The collection covered nearly the entire geographical distribution of the species (Fig. 1A, B). Samples were collected in countries where the species is known to be abundant and/or of commercial value in northern America and where it is classified as invasive or naturalized in Europe. In each population, two or three leaves were sampled from 10 – 20 randomly chosen trees. The leaves were either dried for shipping or frozen in liquid nitrogen. DNA was extracted from 1 cm2of leaf tissue using a modified CTAB protocol (Pairon and Jacquemart, 2005).
Microsatellite and cpDNA typing
Eight nuclear microsatellite loci previously described for P. serotina (Table 2; Pairon et al., 2008) were amplified in all sampled individuals. Polymerase chain reactions (PCR) were performed in a total volume of 15 mL using the methods previously described (Pairon et al., 2006b). PCR pro-ducts were separated on an ABI3100 genetic analyser (Applied Biosystems) and individuals genotyped using GENEMAPPER3.5 (Applied Biosystems).
Five cpDNA loci previously described for P. serotina (Petitpierre et al., 2009) were used on a subsample of ten ran-domly chosen individuals per population (i.e. 230 individuals in the invasive range and 220 in the native range). These markers are derived from non-coding plastid regions and cor-respond either to length variation in polyT/A stretches (ccmp5, trnT-trnL-poly-T and trnT-trnL-poly-A) or to restriction frag-ment length polymorphisms due to a single nucleotidic substi-tution (trnD-trnT-TasI and trnS-trnG-TasI).
Data analyses
Analysis of nuclear microsatellite data.For each microsatellite locus, the total number of alleles, Nei’s unbiased estimate of gene diversity (HS;Nei, 1987), and the inbreeding coefficient
(FIS) were separately estimated for invasive and native
popu-lations using FSTAT 2.9 (Goudet, 1995). To characterize further the genetic structure within loci, Wright’s fixation index (FST) and its analogue based on allele size (RST;
TA B L E1. Geographical location of the Prunus serotina populations sampled in this study together with their basic genetic
information content at eight microsatellite loci
Pop ID Location Country (state) Latitude Longitude N NA NPA
†
A HO H FIS
Native range
CA Gue´nette Canada (Quebec) 46.50 8N 275.25 8W 15 6.63 4 (0) 5.92 0.75 0.74 20.02
IN Bloomington USA (Indiana) 39.19 8N 286.51 8W 15 8.00 4 (1) 6.75 0.62 0.70 0.12
IA1 Ames USA (Iowa) 42.04 8N 293.68 8W 15 8.25 4 (0) 7.17 0.83 0.81 20.02
IA2 New Sharon USA (Iowa) 41.58 8N 292.50 8W 15 6.88 6 (1) 5.92 0.60 0.70 0.14***
MO Springfield USA (Missouri) 37.17 8N 293.33 8W 15 8.00 13 (2) 6.93 0.69 0.75 0.08
NC Otto USA (North Carolina) 35.03 8N 283.48 8W 10 6.50 4 (0) 6.31 0.67 0.76 0.12
NE1 Lincoln USA (Nebraska) 40.81 8N 296.64 8W 14 7.50 16 (2) 6.64 0.66 0.73 0.09
NE2 Omaha USA (Nebraska) 41.26 8N 296.01 8W 15 8.25 14 (1) 7.15 0.75 0.78 0.03
OH1 Ashtabula USA (Ohio) 41.89 8N 280.79 8W 15 7.50 4 (0) 6.58 0.72 0.74 0.03
OH2 Delaware USA (Ohio) 40.37 8N 283.05 8W 14 8.50 14 (1) 7.27 0.73 0.79 0.07
OH3 Ashville USA (Ohio) 39.67 8N 282.93 8W 14 7.75 8 (1) 6.72 0.74 0.73 20.01
OK Broken Bow USA (Oklahoma) 33.97 8N 294.54 8W 13 6.13 13 (3) 5.61 0.64 0.63 20.02
PE1 Behrend USA (Pennsylvania) 42.11 8N 279.98 8W 15 7.38 2 (0) 6.38 0.74 0.76 0.02
PE7 Denton Hill USA (Pennsylvania) 41.77 8N 277.82 8W 15 8.50 12 (0) 7.45 0.82 0.81 0 PE2 Buckaloons USA (Pennsylvania) 41.80 8N 279.25 8W 15 9.38 11 (2) 7.83 0.74 0.76 0.03 PE3 Chapman Dam USA (Pennsylvania) 41.75 8N 279.17 8W 15 7.75 10 (0) 6.66 0.71 0.79 0.1
PE4 Tionesta USA (Pennsylvania) 41.71 8N 278.94 8W 15 8.25 8 (0) 7.27 0.84 0.80 20.05
NY Pawling USA (Pennsylvania) 41.58 8N 273.61 8W 15 7.88 3 (0) 6.74 0.77 0.78 0
SC Charleston USA (South Carolina) 32.75 8N 279.91 8W 15 7.13 12 (1) 6.32 0.68 0.78 0.13*
WI1 Saukville USA (Wisconsin) 43.39 8N 288.02 8W 15 8.13 6 (1) 6.83 0.71 0.74 0.05
WI2 Madison USA (Wisconsin) 43.04 8N 289.44 8W 15 8.63 14 (2) 7.31 0.77 0.77 0.01
WV Spruce Knob USA (West Virginia) 38.71 8N 279.54 8W 14 7.38 5 (0) 6.70 0.82 0.79 20.04
Mean 7.74 – 6.75 0.72 0.76 0.03
Invasive range
BE Lagland Belgium 49.40 8N 05.46 8E 20 7.13 0 (1) 5.69 0.72 0.72 0.007
DA1 Copenhagen Denmark 55.95 8N 12.27 8E 20 6.88 0 5.83 0.74 0.75 0.006
DA2 Tisvilde Denmark 56.05 8N 12.08 8E 14 7.00 0 6.28 0.79 0.74 20.069
DA3 Aarhus Denmark 56.18 8N 9.65 8E 14 6.38 0 5.82 0.71 0.73 0.04
FR1 Fontainebleau France 48.38 8N 2.75 8E 20 4.63 0 (1) 4.14 0.57 0.59 0.035
FR2 Alsace France 48.72 8N 7.68 8E 16 6.50 0 5.65 0.66 0.72 0.09
FR3 Compie`gne France 49.37 8N 2.88 8E 20 7.25 0 (1) 6.08 0.74 0.74 20.004
FR4 Bordeaux France 44.13 8N 0.90 8W 20 6.38 0 (2) 5.29 0.71 0.70 20.023
GE1 Potsdam S Germany 52.73 8N 12.30 8E 20 7.5 0 (1) 5.46 0.76 0.75 0.008
GE2 Potsdam T Germany 52.47 8N 13.22 8E 20 6.38 0 6.16 0.71 0.72 20.003
GE3 Neuwig Germany 50.43 8N 7.63 8E 20 7.13 0 5.94 0.72 0.72 0.008
GE4 Mannheim N Germany 49.65 8N 8.53 8E 20 7.25 0 5.95 0.69 0.74 0.068
GE5 Karlsruhe Germany 48.97 8N 8.37 8E 20 7.13 0 5.87 0.73 0.70 20.042
GE6 Saarbrucken Germany 49.30 8N 7.23 8E 19 6.75 0 5.78 0.68 0.74 0.08
GE7 Cuxhaven Germany 53.82 8N 8.63 8E 20 6.63 0 5.67 0.69 0.74 0.07
HO1 Arnhem The Netherlands 52.00 8N 5.83 8E 20 7.25 0 5.82 0.73 0.71 20.014
HO2 Amsterdam The Netherlands 52.35 8N 4.55 8E 19 6.63 0 5.54 0.63 0.71 0.123
IT1 Ticino Italy 45.52 8N 8.72 8E 20 4.50 0 4.08 0.66 0.64 20.018
IT2 Udine Italy 46.12 8N 13.17 8E 20 3.13 0 2.98 0.60 0.55 20.088
PO1 Poznan Poland 52.17 8N 17.15 8E 20 5.38 0 4.76 0.67 0.68 0.015
PO2 Wroclaw Poland 51.37 8N 16.55 8E 20 5.00 0 4.49 0.69 0.61 20.119
UK1 Shakelford UK 51.18 8N 0.65 8W 20 4.63 0 3.91 0.59 0.59 0.006
UK2 Bagshot UK 51.38 8N 0.73 8W 20 5.50 0 (2) 5.03 0.76 0.75 20.011
Mean 6.36 – 5.31 0.69 0.70 0.01
N is the number of individuals sampled per population, NAis the mean number of alleles per locus, A is the mean allelic richness per locus (based on n ¼ 10) and NPAis the number of private alleles. Estimates of observed heterozygosity (HO), gene diversity (H ) and inbreeding coefficient (FIS) are also given.
* 0.01 , P , 0.05, *** P , 0.001
†For the private alleles, the first number is the number of alleles private to a given range (invasive or native) in the considered population, while the number in parenthesis is the number of alleles that are private to the considered population in its range.
AA A A A A A A B OK WV CA MO OH3 WI2 IN OH2 PE2 NY SC PE1 OH1 WI1 IA2 PE3 PE4 PE7 NE1 IA1 NE2 A NC Native distribution of
P. serotina var. serotina A BE HO2 GE7 DA2 UK1 IT2 IT1 FR1 FR4 UK2 FR3 FR2 HO1 DA1 PO1 PO2 GE4 GE2 GE5 GE6 GE3 DA3 GE1
Density of georeferrenced occurrences of
P. serotina in a 100 km radius
Countries where P. serotina is widely naturalized Countries where P. serotina is not signaled
Countries where P. serotina is locally naturalized
Countries where P. serotina is signaled but no problematic
B C ss4 ss1 ss5 ss2 ss3 ss6 ss4 ss1 Native range Invaded range A A A A A A AAA C 1 2 3 45 8 8 8 8 6 6 8 8 7 8 8 8 8 8 8 8 8 8 A A 0 350 700 1400 km 0 350 700 1400 km
FI G. 1. Geographical distribution of nuclear and chloroplast genetic variation in Prunus serotina. (A) Native range with sampling locations. The hatched area indicates the natural occurrence of P. serotina var. serotina in the wild. Dots on the map represent the three genetic clusters defined by BAPSbased on nuclear
microsatellite data. These dots are labelled from A to C according to the cluster they belong to. Pie charts represent the relative proportions of cpDNA haplotypes in a population, as indicated in (C). (B) Invasive range with sampling locations. Dots on the map represent the eight genetic clusters defined by BAPSbased on the
nuclear microsatellite data. The dots are numbered from one to eight according to the cluster they belong to. (C) Median-joining network of cpDNA haplotypes. The size of the circles represents the relative frequency of the haplotypes. The biggest size corresponds to haplotypes with a frequency .10 %, the intermediate size corresponds to haplotypes with a frequency between 1 and 10 % and the smallest size corresponds to haplotypes with a frequency ,1 %. Steps between
Michalakis and Excoffier, 1996) were computed using SPAGEDi 1.2 (Hardy and Vekemans, 2002). To detect whether FSTand
RST were significantly different from 0, SPAGEDi was also
used to perform a 10 000-permutations test on alleles. FST
values cannot be higher than the homozygosity level, 1 – HS (Hedrick, 2005). F′ST, an estimator unconstrained by
genetic diversity, was therefore calculated as FST/1 – HS.
10 000 randomizations tests were performed to assess the differences of genetic diversity and differentiation between the native and invaded range. The partitioning of the gene diversity at different geographic scales (i.e. within population, among populations and between invasive and native ranges) was also measured using an AMOVA with ARLEQUIN 3.1.1 (Excoffier et al., 2005).
To characterize the genetic diversity at the population level over the eight loci, the observed heterozygosity (HO) was
cal-culated with GENALEX6.0 (Peakall and Smouse, 2006), as well as the allelic richness (A), Nei’s unbiased estimate of gene diversity (HS;Nei, 1987) and the inbreeding coefficient (FIS)
with FSTAT. Significance of the inbreeding coefficient was assessed with FSTAT using permutations. Hardy – Weinberg equilibrium was tested using GENEPOP 3.4 (Rousset, 2008). The number of private alleles was computed for each popu-lation in the native range.
Introduced populations are often subject to founder effects that result in reduced genetic diversity (i.e. a bottleneck). To assess the decrease in diversity from native to invasive popu-lations, three different methods were used. First, the values of allelic richness and gene diversity averaged over the whole invasive and native ranges were compared, postulating that a genetic bottleneck should decrease the number of alleles present in the invasive range. Differences in A and HS were assessed using a paired t-test in the SAS software
version 8.2 (SAS Institute, 1999). Secondly, BOTTLENECK1.2 (Cornuet and Luikart, 1996) and the Wilcoxon test for hetero-zygote excess were used under the two-phase model and the stepwise mutation model. The Wilcoxon test is preferable when few loci are used (Cornuet and Luikart, 1996; Cornuet et al., 1999). Thirdly, the allele frequencies were plotted for each invasive population based on the method described in Luikart et al. (1998). This method postulates that recent bottle-necks should cause a decrease in rare alleles, creating a charac-teristic mode-shift distortion in the distribution of allele frequencies.
The population genetic structure at the geographic level is shaped by different forces such as genetic drift and the possi-bility of inter-population gene exchanges. The characterization of genetic relationships between populations can be informa-tive about their histories. The relatedness among populations was first assessed by computing F-statistics (Weir and Cockerham, 1984) and Nei’s standard genetic distance (D; Nei, 1972) between pairs of invasive populations. A signifi-cance test of 10 000 permutations of the estimates of FST
and D obtained was performed using GENETIX 4.052 (Belkhir et al., 2000). The genetic structure in the invasive range was further described, together with that of the native range, by using the Bayesian clustering method implemented in BAPS 4.14 (Corander and Marttinen, 2006). BAPSuses an analytical integration strategy combined with stochastic search methods to infer the number of clusters K and was run using both
T AB LE 2. Genetic div ersity detected a t eight micr osa tellite loci in invasiv e and na tiv e Prunus ser otina Al lele si ze rang e T ot al no. of alleles HS FIS FST F ′ ST Lo cus Inv asiv e N ativ e Inv asiv e (n ¼ 442) Na tiv e (n ¼ 321) Inv asiv e N ativ e Invasiv e N ativ e Invasiv e N ativ e W h ole da taset Invasiv e N ativ e Whol e d ataset UDP9 6-005 82 – 8 8 8 2 – 96 4 7 0 . 57 0 . 60 0 . 02 5 ns 0 . 09 ns 0 . 06* 0 . 09* 0 . 09* 0 . 16* 0 . 24* 0 . 24* UDP9 6-025 10 9 – 137 10 8 – 137 13 18 0 . 78 0 . 84 2 0 . 00 8 ns 0 . 07 * 0 . 09* 0 . 03* 0 . 08* 0 . 41* 0 . 20* 0 . 42* UDP9 6-405 10 9 – 123 10 9 – 123 6 8 0 . 51 0 . 63 0 . 11 9 ns 0 . 01 ns 0 . 15* 0 . 09* 0 . 14* 0 . 32* 0 . 26* 0 . 32* UCD-CH14 13 0 – 168 12 4 – 182 19 29 0 . 81 0 . 90 2 0 . 00 2 ns 0 . 02 ns 0 . 10* 0 . 04* 0 . 09* 0 . 56* 0 . 47* 0 . 60* PceGA34 13 5 – 175 13 1 – 175 19 24 0 . 80 0 . 89 2 0 . 03 0 ns 0 . 01 ns 0 . 08* 0 . 05* 0 . 08* 0 . 38* 0 . 45* 0 . 53* Pchg ms2 12 7 – 159 12 5 – 163 11 18 0 . 69 0 . 55 0 . 00 7 ns 2 0 . 02 ns 0 . 09* 0 . 13* 0 . 12* 0 . 30* 0 . 29* 0 . 32* Pchp gms3-1 16 8 – 198 16 6 – 198 11 12 0 . 65 0 . 79 2 0 . 01 2 ns 0 . 05 ns 0 . 10* 0 . 05* 0 . 10* 0 . 30* 0 . 23* 0 . 33* Pchp gms3-2 20 0 – 238 20 0 – 238 15 18 0 . 80 0 . 86 0 . 01 1 ns 0 . 08 ns 0 . 07* 0 . 05* 0 . 06* 0 . 35* 0 . 32* 0 . 37* Mean 0 . 70 0 . 76 0 . 01 ns 0 . 03 ns 0 . 09* 0 . 06* 0 . 09* 0 . 35* 0 . 31* 0 . 39* The Nei’s gene div ersi ty per locu s (H S ) and W right’ s inbr eedin g coefficient (F IS ) w er e av er aged o v er the na tiv e and invasi v e popu la tions. Th e W right’s fixa tion inde x (F ST ) and an analo gue uncon st rained by gen e div ersity (F ′ ST ) w er e av er aged o v er the na tiv e and invasiv e popula tions or the whole da taset. ns, not sign ificant; *, P , 0 . 001.
genotype and sample group information (group mode). The maximum number of populations ( prior information) was set to 40.
The genetic clusters defined by BAPS in the native range were then used as reference populations in GENECLASS2 (Piry et al., 2004) to assign individuals from invasive populations. The frequency-based method of Paetkau et al. (1995) and Monte-Carlo resampling was used to compute assignment probabilities with the simulation algorithm of Piry et al. (2004). To gain insight into the source of the clusters defined by BAPSin the invasive range, the proportion of indi-viduals assigned to the different clusters of the native range was calculated within each cluster of the invasive range.
Analysis of cpDNA data.To characterize the genetic diversity of the chloroplastic genome, the allelic richness (A), cpDNA hap-lotype diversity within populations (HS), total cpDNA
haplo-type diversity (HT) and the level of genetic differentiation
(FST) within both ranges were calculated using FSTAT. Ten
thousand randomizations tests were performed to assess the differences of genetic diversity and differentiation between the native and invaded range. Relationships among cpDNA haplotypes were visualized by constructing a statistical median-joining network implemented in NETWORK 4.112 (Bandelt et al., 1999).
Haplotypes were then plotted on maps to illustrate their geo-graphical distribution and the putative introduction sources. Their percentage distribution was also estimated within both ranges.
R E S U LT S
Genetic diversity
Within microsatellite loci. All eight nuclear microsatellite loci were polymorphic in each population, and a total number of 134 alleles was observed on the whole sample (Table 2). Ninety-eight alleles were observed in invasive populations, while all of them were detected in native populations (Table 2). Thirty-six alleles (27 %) were private to the native range. Overall gene diversity (HS) was also higher in the
native range at the majority of loci (Table 2). FIS ranged
from 20.02 to 0.09 and a significant deviation from Hardy – Weinberg equilibrium was found at locus UDP96-025 but only in the native range. FSTshowed an overall differentiation
of 0.06 in the native range and 0.09 in the invasive range. RST
values were not significantly different from FSTand were thus
not shown. F′ST values also showed that the overall
differen-tiation in the native range (0.31) was lower than in the invasive range (0.35; Table 2) but this difference was not significant. The AMOVA analysis indicated that genetic diversity was mainly partitioned within populations (89 + 2.9 %), then among populations (8 + 0.3 %), and then between the native and invasive ranges (3 + 0.1 %).
Chloroplastic DNA haplotypes. Considering allelic combi-nations at the five cpDNA loci, six haplotypes were detected in the whole dataset. Both native and invasive ranges shared four haplotypes (Table 3). Haplotype ss2 was private to the native range and only found in the western part of the United States (Fig. 1A). On the other hand, haplotype ss6
was private to the invasive range and exclusively found in northern Italy (Fig.1B). Haplotypes ss1 and ss4 were by far the most abundant (Table 3), and displayed a central position in the median-joining network (Fig. 1C). This network also suggests that ss2 and ss3 have derived from ss1 while ss5 and ss6 have derived from ss4. The level of genetic differen-tiation measured by FST averaged 0.40 and 0.52 in native
and invasive populations (Table 4), respectively, but these values were not significantly different (P ¼ 0.49).
Sources of invasive populations
Nuclear microsatellite data. BAPS in group mode found three clusters in the native range (Fig. 1A) and eight clusters in the invasive range (Fig. 1B). The three clusters found in the native range only showed that one cluster (A) was formed by all populations except OK and SC, which accounted for the two remaining groups (B and C, respectively). The genetic structure was more pronounced in the invasive range and the output of the program suggested that the two populations from the United Kingdom (UK1-UK2), populations FR1, FR3 and FR4 (France), and population PO2 (Poland) were all distinct genetic clusters. The two populations from Italy represented another independent cluster, and the last cluster was formed by the remaining populations (i.e. populations from Denmark, Germany, The Netherlands and Belgium and populations FR2 and PO1). When the individual genotypes from the invasive range were used to assign each individual to one of the three clusters of the native range, very few trees were assigned to clusters B and C, and the vast majority was assigned to the main cluster (A) in the native range. When the clusters found in the invasive range were reconstructed and the proportion of individuals within each cluster that were
TA B L E3. Compared occurrences of the six cpDNA haplotypes
(with their frequency given in parenthesis) in both native and invasive ranges of Prunus serotina
Haplotype no. Native range Invasive range
ss1 96 (43.8 %) 9 (3.9 %) ss2 8 (3.7 %) – ss3 2 (0.9 %) 34 (14.8 %) ss4 112 (51.1 %) 164 (71.3 %) ss5 1 (0.5 %) 3 (1.3 %) ss6 – 20 (8.7 %)
TA B L E4. Chloroplast DNA diversity in native and invasive
Prunus serotina populations
Range N NA A HS HT FST
Native 219 5 1.953 0.333 0.547 0.402*
Invasive 230 5 1.651 0.225 0.461 0.519*
Haplotypes were defined by the combination of alleles at five loci. Number of individuals (N ), number of haplotypes (NA), mean allelic richness per population (A), haplotype diversity within populations (HS), total haplotype diversity (HT) and the level of genetic differentiation (FST) are given for each range.
assigned to the three main clusters of the native range were looked at, it was noted that 83 – 95 % of the individuals within clusters 2, 3, 5, 6, 7 and 8 were assigned to the main cluster (A) in the native range (Table 5). Clusters 1 and 4, comprising populations UK2 and FR1, respectively, showed a relatively low proportion of individuals assigned to the main cluster in the native range (65 % and 0 %, respectively), and even to any of the two other native clusters (0 – 15 % and 0 – 50 %, respectively).
Chloroplastic DNA data. cpDNA haplotype distribution gave more information as to which regions in the native range may have acted as source(s) for invasive populations (Table3). This distribution was indeed quite different among the two ranges, and in the invasive area, a much lower pro-portion of the ss1 haplotype was detected, whereas ss3 and ss4 were overrepresented. The vast majority of populations located to the east of the Appalachian Mountains showed a predominant proportion of ss4, as observed in the invasive range, whereas most populations located to the west to the Appalachian Mountains showed a predominant proportion of ss1. Interestingly, ss3 was found in half of the European popu-lations and only detected in two popupopu-lations of the Allegheny plateau in the USA (PE2 and PE4; Fig.1A). Finally, ss2 was not observed in the invasive range and was only found in the western part of the native range.
Partition of genetic diversity within native and invasive ranges Within the native range.The 36 alleles that were private to the native range were relatively well distributed among the native populations (Table 1). Populations PE1 and NY (Pennsylvania) displayed, respectively, only two and three alleles that were not found in the invasive range whereas a maximum of 16 private alleles were found within population NE1 (Nebraska). The number of alleles per locus (NA) was
7.74 in average, ranging from 6.13 to 9.38 per population. The mean allelic richness (A) ranged from 5.61 to 7.83 with a mean of 6.75. Gene diversity (HS) ranged from 0.63 to
0.81 with a mean of 0.76 whereas mean observed heterozygos-ity (HO) ranged from 0.60 to 0.84 among populations with a
mean of 0.72. The inbreeding coefficient (FIS) varied widely
among populations, ranging from 20.05 to 0.14 with a mean of 0.03. Significant deviations from Hardy – Weinberg genotypic proportions associated with positive FIS values
were found for populations IA2 and SC.
Within the invasive range. The mean allelic richness (A) was significantly lower in the invasive range (5.31) than in the native range (6.75) (t ¼ 6.79, P , 0.0001) as expected by the higher number of alleles found in the native range (Table 1). This also resulted in a significantly lower gene diversity in the invasive range (t ¼ 3.75, P ¼ 0.0005; Table1). The number of alleles per population from the inva-sive range varied more widely than in the native range. Populations FR1 (France), IT1, IT2 (Italy), PO1, PO2 (Poland), UK1 (United Kingdom) displayed the lowest amount of genetic diversity. Gene diversity (HS) varied from
0.55 to 0.75 whereas mean observed heterozygosity (HO)
ranged from 0.57 to 0.79 among populations. None of the populations showed significant departure from Hardy – Weinberg proportions and inbreeding coefficients (FIS)
ranged from 20.12 to 0.07, with a mean of 0.01.
Even though some populations were identified as having a low number of alleles and a low gene diversity, only five popu-lations at the margins of the invasive range (IT1, IT2, PO2, DA1 and UK2) were significantly detected as bottlenecked based on the Wilcoxon test under the two-phase model, and only two populations (IT2 and UK2) under the stepwise mutation model. In addition, the graphical method allowed two recently bottlenecked populations (PO1 and PO2) to be detected, with a marked distortion in the distribution of allele frequencies.
A maximum of two cpDNA haplotypes was found per popu-lation in the invasive range, whereas some popupopu-lations of the native range exhibit three haplotypes (Fig. 1A and B). Allelic richness was significantly lower in the invasive range (A ¼ 1.65) than in the native range (A ¼ 1.95) ( permutation test, P ¼ 0.04; Table4). The haplotype diversity within popu-lations (HS) and the total haplotype diversity (HT) were lower
in the invasive than in the native range (Table 4), and this difference was significant for HT ( permutation tests, P ¼
0.02) but not for HS( permutation tests, P ¼ 0.054).
D I S C U S S I O N
The combined use of cpDNA and nuclear microsatellite loci aimed at comparing the genetic diversity of P. serotina var. ser-otina populations between native and invasive ranges and detecting the putative sources of invasive populations. Previous studies have demonstrated the importance of extensive sampling in both ranges (e.g.Novak and Mack, 1993;Eckert et al., 1996;Meekins et al., 2001;Okada et al., 2009) and the usefulness of combining markers with two different modes of inheritance (Williams et al., 2005) to achieve these goals, but as far as is known, the present study is one of the first attempts to answer these questions for a long-lived species.
Sources of invasive P. serotina
The presence of five cpDNA haplotypes in Europe suggests there were multiple introductions of individuals. The
TA B L E5. Proportion of individuals belonging to the eight
clusters from the invasive range (Fig. 1B) assigned to each of the three clusters defined in the native range (Fig.1A)
Clusters in native range Clusters in invasive range A (293) B (13) C (15)
1 (20) 0.65 0.00 0.15 2 (20) 0.95 0.00 0.05 3 (20) 0.95 0.00 0.05 4 (20) 0.00 0.50 0.00 5 (20) 0.95 0.00 0.15 6 (40) 0.83 0.00 0.25 7 (20) 0.95 0.05 0.30 8 (282) 0.90 0.00 0.20
The number of individuals within each cluster (for both native and invasive ranges) is given in parenthesis. Note that individuals can be assigned to more than one cluster.
distribution of haplotypes in the native range can be used to identify putative sources of invasive populations (e.g. Besnard et al., 2007; Hufbauer and Sforza, 2008). First, fre-quencies of haplotypes ss1 and ss4 in both ranges and the dis-tribution of haplotype ss2 all suggest that invasive P. serotina trees mainly originated from the eastern United States. Haplotype ss4 is very frequent in the invasive range (Table 3) and mainly distributed to the east of the Appalachian Mountains whereas ss1 is almost absent in Europe and very frequent in the western part of the native range (Fig.1A). Moreover, ss2 is absent in Europe and only found in the north-western states of Indiana and Nebraska, further suggesting that such populations should not be con-sidered as putative source regions. Secondly, ss3 is overrepre-sented in many invasive populations in Europe suggesting that the Allegheny plateau was an important source of introduction as this haplotype is only found in native populations PE2 and PE4 (where ss4 and ss1 are also present). Prunus serotina is harvested for timber only in the Allegheny plateau (Hough, 1965; Marquis, 1990) and it is not surprising that this area was an important source for introductions to Europe.
No strong geographic differentiation based on nuclear loci in the native range was detected. Only two populations in the southern native range (SC and OK) were distinguished from other populations (Fig. 1A). Their geographic isolation in the native range probably explains their high genetic differ-entiation from the main native cluster A. Human movement of the species for forestry purposes and high seed dispersal in the native range may in part be responsible for the low level of genetic differentiation within the large cluster A. Assignment tests of invasive individuals were not very informative as they only supported the assertion that cluster A was probably involved in most introductions to Europe. Nevertheless, the low number of native-range private alleles in some populations from the Allegheny plateau (e.g. PE1 and NY) again suggests, as cpDNA haplotypes, that this area within cluster A was the main source of introductions to Europe.
Genetic structure and low reduction of gene diversity in the invaded range
The overall genetic structure displayed by microsatellite loci in the present study was more pronounced in the invasive range than in the native range. Populations from Denmark, Germany, The Netherlands and Belgium had higher genetic diversity and lower differentiation among populations, while some populations from the United Kingdom, France and Italy had low allelic richness, low gene diversity and high genetic differentiation among populations. These two different patterns were supported by the output of BAPS as populations from Denmark, Germany, The Netherlands and Belgium were clustered together (cluster 8) whereas the other popu-lations belonged to different separated clusters.
A reduction in the overall number of alleles, allelic richness and heterozygosity was found at microsatellite loci in the inva-sive range. Most introductions result in a temporary high reduction in the effective size of populations, and a reduction in diversity within invasive populations should therefore often be observed (Dlugosch and Parker, 2008). This reduction was particularly evident at microsatellite loci as these markers
contain many rare alleles, and are therefore more sensitive to founder effects (Spencer et al., 2000). Despite the observed reduction in allelic diversity, strong evidence of population bot-tlenecks was only observed in some populations at the margins of the invaded range that probably correspond to recent introduc-tions (PO2, DA1, IT1, IT2 and UK2). Indeed, the tests performed look at recent bottleneck events by searching for the allele depletion that should be generated by a sudden decrease of repro-ducing individuals in a newly formed population. Because allele depletion is a complex function that depends on several par-ameters, including the time since a population bottleneck (Cornuet and Luikart, 1996; Prentis et al., 2008; Bucharova and van Kleunen, 2009), populations that had gone through a bot-tleneck several generations ago may not display allelic depletion any more (Dlugosch and Parker, 2008). This points to the impor-tance of the temporal nature of introductions in genetic diversity studies as the particular time of study might affect the ability to detect past bottlenecks (Dlugosch and Parker, 2008).
Despite the reduction in genetic diversity in some margin-ally distributed populations, the invasive range in general pos-sessed high levels of diversity (73 % of allelic richness), with an overall high gene diversity with microsatellite markers and the same number of cpDNA haplotypes in both ranges. The low reduction in genetic diversity can be attributed to multiple introductions with large numbers of individuals and the rela-tively long time since introduction (Bucharova and van Kleunen, 2009).
The initial propagule size and the occurrence of gene flow between established populations seem in the present case to be important factors explaining the genetic patterns observed in the invasive range. All species are likely to experience periods when founding populations are isolated from other conspecific populations, but this relative isolation cannot always be recorded, especially when natural gene flow has been rapidly restored among expanding populations, multiple introductions are geographically close to one another and/or human-mediated seed exchanges occur among introduced populations (Suarez and Tsutsui, 2008). Interestingly, geneti-cally less diverse populations from France, Italy, Poland and England (FR1, IT1 – IT2, PO1 – PO2, UK1 – UK2) are all situ-ated on the margins of the species range in Europe and in countries where no records of massive plantations for forestry purposes are known (Starfinger, 1997). Populations from The Netherlands, Belgium, Germany and Denmark were more diverse and historical records support that propagule number was more important, with several massive introductions of the species for different purposes during the 19th century (Muys et al., 1992; Starfinger, 1997). The relatively high genetic diversity in populations from north-western Europe probably contributed to the adaptive success and the invasive behaviour of P. serotina.
AC K N OW L E D G E M E N T S
This study was supported by the Fonds Spe´ciaux de Recherche (FSR) of the Universite´ catholique de Louvain (UCL), the Belgian Scientific Policy (BelSPo – InPlanBel) and the National Centre of Competence in Research (NCCR) Plant Survival, research programme of the Swiss National Science Foundation. Collection of some samples was partly funded
by grants to K. O. Reinhart from Highlands Biological Station and from the AW Mellon Foundation. We are grateful for the kind collaboration of many foresters and scientists who helped us to collect leaves of Prunus serotina. Samples from France were provided by P. Geldreich, J. M. Ansolabehere, R. Rodriguez and J. Jaminon (Office National des Foreˆts), and G. Decocq (University of Picardie Jules Verne); from Germany by T. Heinken (University of Potsdam), U. Starfinger (Technische University of Berlin) and J. Meyer (Forestrevier Altenwalde); from The Netherlands by W. van der Putten (NIOO-KNAW) and A. Ehrenburg (Amsterdam Water Supply); from Italy by F. Caronni and L. Hildebrand (Parco Lombardo della Valle del Ticino) and S. Del Fabbro; from Denmark by E. Kjaer (Royal Veterinary and Agricultural University), N. P. Revsbech (University of Aarhus) and O. Raspe´ (Jardin botanique national de Belgique); from Poland by B. Suszka (Polish Academy of Sciences) and A. Halarewicz (University of Agriculture of Wroclaw); from the USA by G. Meyer (University of Wisconsin – Milwaukee Field Station), A. K. Buthod (Robert Bebb Herbarium, The University of Oklahoma), D. A. Lewis (Ada Hayden Herbarium, Iowa State University), L. Iverson (Northern Research Station, USDA Forest Service), R. B. Kaul and D. M. Sutherland (Bessey Herbarium University of Nebraska), T. S. Cochrane (University of Wisconsin – Madison Herbarium), K. O. Reinhart (United States Department of Agriculture, Agricultural Research Service), R. A. Klips (Ohio State University) and L. M. Bowe (Ozarks Regional Herbarium, Missouri State University). We thank L. Dhondt, C. Noe¨l and D. Savova-Bianchi for technical help in the laboratory. We would like to thank O. Hardy, O. Raspe´, U. Starfinger and two anonymous referees for helpful comments and revision of previous versions of the manuscript.
L I T E R AT U R E C I T E D
Bandelt HJ, Forster P, Rohl A. 1999.Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16: 37 – 48. Belkhir K, Borsa P, Goudet J, Chikhi L, Bonhomme F. 2000.GENETIX4.03,
logiciel sous WindowsTM pour la ge´ne´tique des populations. Laboratoire Ge´nome, Populations, Interactions CNRS UMR 5000, Universite´ de Montpellier II, Montpellier, France.
Besnard G, Henry P, Wille L, Cooke D, Chapuis E. 2007.On the origin of the invasive olives (Olea europaea L., Oleaceae). Heredity 99: 608 – 619. Bucharova A, van Kleunen M. 2009.Introduction history and species charac-teristics partly explain naturalization success of North American woody species in Europe. Journal of Ecology 97: 230 – 238.
Carlton JT. 1996.Pattern, process and prediction in marine invasion ecology. Biological Conservation 78: 97 – 106.
Chabrerie O, Verheyen K, Saguez R, Decocq G. 2008. Disentangling relationships between habitat conditions, disturbance history, plant diver-sity, and American black cherry (Prunus serotina Ehrh.) invasion in a European temperate forest. Diversity and Distributions 14: 204 – 212. Corander J, Marttinen P. 2006.Bayesian identification of admixture events
using multi-locus molecular markers. Molecular Ecology 15: 2833 – 2843. Cornuet JM, Luikart G. 1996.Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144: 2001 – 2014.
Cornuet JM, Piry S, Luikart G, Estoup A, Solignac M. 1999.New methods employing multilocus genotypes to select or exclude populations as origins of individuals. Genetics 153: 1989 – 2000.
DAISIE. 2009. Handbook of alien species in Europe. Knoxville, TN: Springer.
Devos N, Tyteca D, Raspe´ O, Wesselingh RA, Jacquemart AL. 2003. Pattern of chloroplast diversity among western European Dactylorhiza species (Orchidaceae). Plant Systematics and Evolution 243: 85 – 97. Dlugosch KM, Parker IM. 2008. Founding events in species invasions:
genetic variation, adaptive evolution, and the role of multiple introduc-tions. Molecular Ecology 17: 431 – 449.
Eckert CG, Manicacci D, Barrett SCH. 1996.Genetic drift and founder effect in native versus introduced populations of an invading plant, Lythrum salicaria (Lythraceae). Evolution 50: 1512 – 1519.
Excoffier L, Laval G, Schneider S. 2005.ARLEQUINver. 3.0: an integrated
software package for population genetics data analysis. Evolutionary Bioinformatics Online 1: 47 – 50.
Facon B, Pointier JP, Jarne P, Sarda V, David P. 2008.High genetic var-iance in life-history strategies within invasive populations by way of mul-tiple introductions. Current Biology 18: 363 – 367.
Forman J. 2003.The introduction of American plant species into Europe: issues and consequences. In: Child L, Brock J, Brundu G, et al eds. Plant invasions: ecological threats and management solutions. Backhuys: Leiden, 17 – 33.
Gaskin JF, Zhang DY, Bon MC. 2005. Invasion of Lepidium draba (Brassicaceae) in the western United States: distributions and origins of chloroplast DNA haplotypes. Molecular Ecology 14: 2331 – 2341. Goudet J. 1995. FSTAT (Version 1.2): a computer program to calculate
F-statistics. Journal of Heredity 86: 485 – 486.
Grapputo A, Boman S, Lindstrom L, Lyytinen A, Mappes J. 2005.The voyage of an invasive species across continents: genetic diversity of North American and European Colorado potato beetle populations. Molecular Ecology 14: 4207 – 4219.
Hardy OJ, Vekemans X. 2002.SPAGEDI: a versatile computer program to analyse spatial genetic structure at the individual or population levels. Molecular Ecology Notes 2: 618 – 620.
Hedrick PW. 2005.A standardized genetic differentiation measure. Evolution 59: 1633 – 1638.
Henry P, Le Lay G, Goudet J, Guisan A, Jahodova S, Besnard G. 2009. Reduced genetic diversity, increased isolation and multiple introductions of invasive Giant Hogweed in the western Swiss Alps. Molecular Ecology 18: 2819 – 2831.
Hough AF. 1965.Black cherry (Prunus serotina Ehrh.). In: Fowells HA. ed. Silvics of forest trees of the United States. Agriculture Handbook No. 271. Washington DC: US Department of Agriculture, 539 – 545. Hufbauer RA, Sforza R. 2008. Multiple introductions of two invasive
Centaurea taxa inferred from cpDNA haplotypes. Diversity and Distribution 14: 252 – 261.
Le Roux JJ, Wieczorek AM, Meyer JY. 2008.Genetic diversity and structure of the invasive tree Miconia calvescens in Pacific islands. Diversity and Distributions 14: 935 – 948.
Lockwood JL, Cassey P, Blackburn T. 2005.The role of propagule pressure in explaining species invasions. Trends in Ecology & Evolution 20: 223 – 228.
Luikart G, Allendorf FW, Cornuet JM, Sherwin WB. 1998.Distortion of allele frequency distributions provides a test for recent population bottle-necks. Journal of Heredity 89: 238 – 247.
McCauley DE, Smith RA, Lisenby JD, Hsieh C. 2003. The hierarchical spatial distribution of chloroplast DNA polymorphism across the intro-duced range of Silene vulgaris. Molecular Ecology 12: 3227 – 3235. McVaugh R. 1951.A revision of the north American black cherries (Prunus
serotina Ehrh., and relatives). Brittonia 7: 279 – 315.
Marquis D. 1990. Prunus serotinaEhrh. black cherry. In: Burns RM, Honkala BH. eds. Silvics of North America. Vol. 2. Hardwoods. Washington, DC: US Department of Agriculture, Forest Service, 594 – 604.
Meekins JF, Ballard HE, McCarthy BC. 2001.Genetic variation and mol-ecular biogeography of a North American invasive plant species (Alliaria petiolata, Brassicaceae). International Journal of Plant Sciences 162: 161 – 169.
Michalakis Y, Excoffier L. 1996.A generic estimation of population subdivi-sion using distances between alleles with special reference for microsatel-lite loci. Genetics 142: 1061 – 1064.
Muys B, Maddelein D, Lust N. 1992.Ecology, practice and policy of black cherry (Prunus serotina Ehrh.) management in Belgium. Silva Gandavensis 27: 28 – 45.
Nei M. 1972.Genetic distance between populations. American Naturalist 106: 283 – 292.
Nei M. 1987. Molecular evolutionary genetics. New York, NY: Columbia University Press.
Neuffer B, Hurka H. 1999.Colonization history and introduction dynamics of Capsella bursa-pastoris (Brassicaceae) in North America: isozymes and quantitative traits. Molecular Ecology 8: 1667 – 1681.
Novak SJ, Mack RN. 1993.Genetic variation in Bromus tectorum (Poaceae): comparison between native and introduced populations. Heredity 71: 167 – 176.
Novak SJ, Mack RN. 2001.Tracing plant introduction and spread: genetic evidence from Bromus tectorum (Cheatgrass). BioScience 51: 114 – 122. Okada M, Lyle M, Jasieniuk M. 2009.Inferring the introduction history of the invasive apomictic grass Cortaderia jubata using microsatellite markers. Diversity and Distributions 15: 148 – 157.
Paetkau D, Calvert W, Stirling I, Strobeck C. 1995.Microsatellite analysis of population structure in Canadian polar bears. Molecular Ecology 4: 347 – 354.
Pairon MC. 2007. Ecology and population genetics of an invasive forest tree species: Prunus serotina Ehrh. PhD Dissertation, Universite´ catholique de Louvain.
Pairon MC, Jacquemart AL. 2005.Disomic segregation of microsatellites in the tetraploid Prunus serotina Ehrh. (Rosaceae). Journal of the American Society for Horticultural Science 130: 729 – 734.
Pairon M, Chabrerie O, Mainer Casado C, Jacquemart AL. 2006a.Sexual regeneration traits linked to black cherry (Prunus serotina Ehrh.) inva-siveness. Acta Oecologica 30: 238 – 247.
Pairon MC, Jonard M, Jacquemart AL. 2006b.Modeling seed dispersal of black cherry, an invasive forest tree: how microsatellites may help? Canadian Journal of Forest Research 36: 1385 – 1394.
Pairon MC, Potter D, Jacquemart AL. 2008.Detection and characterization of genome-specific microsatellite markers in the allotetraploid Prunus serotina. Journal of the American Society for Horticultural Science 133: 390 – 395.
Peakall R, Smouse PE. 2006. GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6: 288 – 295.
Petit RJ, Hampe A. 2006.Some evolutionary consequences of being a tree. Annual Review of Ecology, Evolution, and Systematics 37: 187 – 214. Petit RJ, Kremer A, Wagner DB. 1993.Finite island model for organelle and
nuclear genes in plants. Heredity 71: 630 – 641.
Petit RJ, Bialozyt R, Garnier-Gere P, Hampe A. 2004.Ecology and gen-etics of tree invasions: from recent introductions to Quaternary migrations. Forest Ecology and Management 197: 117 – 137.
Petitpierre B, Pairon M, Broennimann O, Jacquemart AL, Guisan A, Besnard G. 2009.Plastid DNA polymorphisms in Prunus serotina var. serotina (Rosaceae), a North American tree invading Europe. European Journal of Forest Research 128: 431 – 436.
Pimentel D, Zuniga R, Morrison D. 2005.Update on the environmental and economic costs associated with alien-invasive species in the United States. Ecological Economics 52: 273 – 288.
Piry S, Alapetite A, Cornuet JM, Paetkau D, Baudouin L, Estoup A. 2004. GENECLASS2: a software for genetic assignment and first-generation migrant detection. Journal of Heredity 95: 536 – 539.
Prentis PJ, Wilson JRU, Dormontt EE, Richardson DM, Lowe AJ. 2008. Adaptive evolution in invasive species. Trends in Plant Science 13: 288 – 294.
Prentis PJ, Sigg DP, Raghu S, Dhileepan K, Pavasovic A, Lowe AJ. 2009. Understanding invasion history: genetic structure and diversity of two globally invasive plants and implications for their management. Diversity and Distributions 15: 822 – 830.
Rousset F. 2008.GENEPOP′007: A complete re-implementation of the GENEPOP
software for Windows and Linux. Molecular Ecology Resources 8: 103 – 106.
SAS Institute. 1999. SAS User’s guide, version 8. Cary, NC: SAS Institute Inc.
Schlaepfer DR, Edwards PJ, Widmer A, Billeter R. 2008.Phylogeography of native ploidy levels and invasive tetraploids of Solidago gigantea. Molecular Ecology 17: 5245 – 5256.
Simberloff D. 2003.Confronting introduced species: a form of xenophobia? Biological Invasions 5: 179 – 192.
Spencer CC, Neigel JE, Leberg PL. 2000.Experimental evaluation of the usefulness of microsatellite DNA for detecting demographic bottlenecks. Molecular Ecology 9: 1517 – 1528.
Starfinger U. 1997. Introduction and naturalization of Prunus serotina in Central Europe. In: Brock J, Wade M, Pysek P, Green D. eds. Plant inva-sions: studies from North America and Europe. Leiden: Backhuys Publishers, 161 – 171.
Starfinger U, Kowarik I, Rode M, Schepker H. 2003.From desirable orna-mental plant to pest to accepted addition to the flora? The perception of an alien tree species through centuries. Biological Invasions 5: 323 – 335. Suarez AV, Tsutsui ND. 2008.The evolutionary consequences of biological
invasions. Molecular Ecology 17: 351 – 360.
Theoharides KA, Dukes JS. 2007. Plant invasion across space and time: factors affecting nonindigenous species success during four stages of invasion. New Phytologist 176: 256 – 273.
Vanhellemont M, Verheyen K, De Keersmaeker L, Vandekerkhove K, Hermy M. 2009.Does Prunus serotina acts as an aggressive invader in areas with a low propagule pressure? Biological Invasions 11: 1451 – 1462.
Vanhellemont M, Wauters L, Baeten L, et al. 2010. Prunus serotina unlashed: invader dominance after 70 years of forest development. Biological Invasions 12: doi 10.1007/s10530-009-9529-x.
Verheyen K, Vanhellemont M, Stock T, Hermy M. 2007.Predicting patterns of invasion by black cherry (Prunus serotina Ehrh.) in Flanders (Belgium) and its impact on the forest understorey community. Diversity and Distributions 13: 487 – 497.
Weir BS, Cockerham CC. 1984.Estimating F-statistics for the analysis of population structure. Evolution 38: 1358 – 1370.
Wilcove DS, Rothstein D, Dubow J, Phillips A, Losos E. 1998.Quantifying threats to imperiled species in the United States. BioScience 48: 607 – 615. Williams DA, Overholt WA, Cuda JP, Hughes CR. 2005.Chloroplast and microsatellite DNA diversities reveal the introduction history of Brazilian peppertree (Schinus terebinthifolius) in Florida. Molecular Ecology 14: 3643 – 3656.
Wilson JRU, Dormontt EE, Prentis PJ, Lowe AJ, Richardson DM. 2009. Something on the way you move: dispersal pathways affect invasion success. Trends in Ecology & Evolution 24: 136 – 144.
With K. 2002. The landscape ecology of invasive spread. Conservation Biology 16: 1192 – 1203.
